Virtually all breast cancer (BC) related deaths result from metastatic burden rather than the primary tumor. It has long been thought that primary tumors are heterogeneous, with only some cells having the capacity to metastasize. Our work suggests that this view is oversimplified, and that non-metastatic cells can become more metastatic due to paracrine-mediated signaling from other cancer cells. Understanding this mechanism will uncover new strategies to inhibit metastatic disease. To understand crosstalk between cells of varying metastatic potential in heterogeneous tumors, we examined whether metastatic cells expressing the EMT-inducing transcription factors (TFs), Twist1 and Snail1, can promote metastatic properties in intrinsically non-metastatic cells. Our data show that, in a manner dependent on a third EMT TF, Six1, the EMT TFs can non-cell autonomously enhance metastatic properties. Mechanistically, we show that cells expressing Six1 mediate paracrine activation of GLI-mediated signaling in intrinsically non-metastatic cells, leading to increased aggressiveness and metastasis of these cells. Importantly, paracrine mediated activation of GLI can occur through non-canonical pathways independent of Hedgehog (Hh) ligands and Smoothened (SMO)-mediated Hedgehog signaling. Our strong data lead us to hypothesize that SMO inhibitors, which are in clinical trials, will fail in a large percentage of BC patients. We hypothesize that targeting GLI directly will be effective in a broader range of breast tumors (encompassing subtypes that show features of EMT), and that determining the molecular mechanism of non-canonical paracrine GLI activation will lead to additional therapies that are effective in a broad range of BCs. Further, we hypothesize that targeting the central mediator, Six1, using our novel small molecule inhibitor, will be efficacious in the broadest range of BCs due to its cell and non-cell autonomous role in tumor progression. To test these related hypotheses, we will first determine the paracrine mechanism by which EMT/metastatic cells increase the aggressiveness of intrinsically non-metastatic cells, focusing on VEGF-C and its downstream signaling as key mediators. We will then perform a large scale head-to-head comparison of inhibitors of EMT/metastatic and non-metastatic cell crosstalk in breast patient derived xenograft (PDX) models that encompass various breast cancer subtypes and are metastatic, to determine which inhibitors are efficacious in a broader range of tumors, and to characterize the types of tumors that will respond to specific inhibitors. Our work may provide a partial explanation for why SMO inhibitors have been ineffective as single agents in tumors such as BCs. Further, by uncovering a new mechanism by which EMT promotes metastasis, and by deciphering the molecular components of that mechanism, we may not only provide an explanation for the controversy in the field, but may more importantly identify highly efficacious means to target the many BCs in which a percentage of cells have undergone an EMT.